4.2 The theoretical phase
Theoretically, to ensure the implementation of the planned change, Firm A set the following theoretical strategies to support the planned transformation. Figure 1 [Figure omitted. See Article Image.] depicts Firm A's theoretical ideas in developing a learning organization. There were four stages.
Stage 1. Leaders from each department were required to attend the study groups, which we called the "virtual hatchery" in this study, to receive the new values that the organization aimed at. Inside the hatchery, through communication and interaction with other leaders, managers started to have a more thorough understanding about other departments' business, job content, and work procedures; so that leaders detected the errors they had, corrected their own cognitive maps, and went back to their departments to adjust and correct the departmental working procedures. This is a learning cycle from tacit knowledge to explicit knowledge. Leaders learned new knowledge and corrected errors in their cognitive systems. And when they returned to their own departments, leaders intended to make decisions with the new tacit knowledge they had learned from the virtual hatchery. Thus a cycle of tacit knowledge to explicit knowledge was completed, and also a cycle of knowledge distribution from the individual level to the departmental/group level. The learning cycle was like an "8" or an infinite "∞", which had no stops and resembled "double loops". Leaders practiced the newly learned beliefs and values in their departments, and got feedback from employees on possible problems, issues, and questions. These questions were brought to the hatchery as "errors" for adjusting the core values inside the virtual hatchery.
At this stage, much of the knowledge was shared inside the hatchery and transformed into individuals' subconsciousness as intuitions and became part of tacit knowledge. When leaders brought back the learned tacit knowledge to their own departments, leaders "interpreted" the tacit knowledge they had learned into explicit knowledge when making departmental decisions. Thus, new values were integrated with the departmental norms, values, and routines gradually. In other words, the intuiting and interpreting of the 4i model ([7] Crossan et al. , 1999) happened in stage 1, and integration happened in stage 2.
Stage 2. In stage 1 the detected errors and adjusted values became initiates to form organizational strategies, norms, and core values. These new values then not only happened at a departmental level, but also at an organizational level, when leaders of each department were required to get together in meetings to make decisions on an organizational level - the tacit knowledge again became explicit knowledge in guiding managers' decision-making during meetings. Then the departmental level learning is transformed to the organizational level. When the newly built knowledge body expanded itself from the departmental level to the organizational level, the organizational norms and core values were adjusted to match with organizational goals and to increase organizational competitiveness. It was also the process from integrating to institution of the learning organization as depicted in the 4i model ([7] Crossan et al. , 1999).
Stage 1 and stage 2's learning were combined as double-loop learning. That was because the main attendances in the study groups were all leaders or key gatekeepers in each department. The leaderships' advantages were used at this point. However, it was usually harder to get managers to change their values, beliefs, etc. So, inside the virtual hatchery, a more soft focus method, which was the adult learning theory, was applied. Every manager/leader was scheduled to host a one-day program with the activities he preferred, or to guide a book study sometimes. And at the beginning or the end of the learning sections, each participant was required to share his feelings for or not for the jobs. This was so that employees had a learning context with less stress and better focus on the learning issues. This learning context also encouraged participants to practice new skills, to look at the issues that had not had attention paid to them before, to communicate with managers from other departments, and to coordinate with others. Gradually, the virtual hatchery became a place that leaders could relax and share information with others with more objective attitudes, as the study group was away from departmental influences.
Stage 3. Making leaders learn was a good idea for knowledge distribution as they were the guides and decision-makers for each department. However, to speed up the construction of a learning organization, single-loop learning had to be applied to every member inside the organization. The final plan was to put all members who were not leaders into the hatchery, and to routinely distribute newly built values and knowledge to organizational members. When the non-leader members were put into the hatchery, members started to use their newly learned knowledge to compare with their existing cognitive map, detect the errors, and make adjustments. The procedure should be smoother, as the leaders had also been "hatched" before. So that, at this stage, Firm A expanded the size and numbers of hatcheries for "hatching" more members who were non-leaders.